Color Distortion-Aware Exposure Fusion

This application claims priority under 35 U.S.C. § 119 (e) to U.S. Provisional Patent Application No. 63/747,596 filed on Jan. 21, 2025, which is hereby incorporated by reference in its entirety.

This disclosure relates generally to image processing systems and methods. More specifically, this disclosure relates to color distortion-aware exposure fusion.

Many mobile electronic devices, such as smartphones and tablet computers, include cameras that can be used to capture still and video images. These types of devices often include image processing pipelines that perform a number of operations in sequence to generate images of scenes. For example, an image processing pipeline may perform operations that combine multiple image frames of a scene into a combined image and process the combined image to produce a final image of the scene.

This disclosure relates to color distortion-aware exposure fusion.

In a first embodiment, a method includes obtaining, using at least one processing device of an electronic device, multiple image frames of a scene, where each image frame includes a luma channel and chroma channels. The method also includes blending, using the at least one processing device, the image frames to generate a blended image. Blending the image frames includes determining first weights to be applied when blending the luma channels of the image frames and separate second weights to be applied when blending the chroma channels of the image frames; generating a fused luma image using the luma channels of the image frames and the first weights and separately generating at least one fused chroma image using the chroma channels of the image frames and the second weights; and combining the fused luma image and the at least one fused chroma image.

In a second embodiment, an electronic device includes at least one processing device configured to obtain multiple image frames of a scene, where each image frame includes a luma channel and chroma channels. The at least one processing device is also configured to blend the image frames to generate a blended image. To blend the image frames, the at least one processing device is configured to determine first weights to be applied when blending the luma channels of the image frames and separate second weights to be applied when blending the chroma channels of the image frames; generate a fused luma image using the luma channels of the image frames and the first weights and separately generate at least one fused chroma image using the chroma channels of the image frames and the second weights; and combine the fused luma image and the at least one fused chroma image.

In a third embodiment, a non-transitory machine readable medium contains instructions that when executed cause at least one processor of an electronic device to obtain multiple image frames of a scene, where each image frame includes a luma channel and chroma channels. The non-transitory machine readable medium also contains instructions that when executed cause the at least one processor blend the image frames to generate a blended image. The instructions that when executed cause the at least one processor to blend the image frames include instructions that when executed cause the at least one processor to determine first weights to be applied when blending the luma channels of the image frames and separate second weights to be applied when blending the chroma channels of the image frames; generate a fused luma image using the luma channels of the image frames and the first weights and separately generate at least one fused chroma image using the chroma channels of the image frames and the second weights; and combine the fused luma image and the at least one fused chroma image.

Any one or any combination of the following features may be used with the first, second, or third embodiment.

The first weights and the second weights may be determined by determining the first weights based on the image frames and modifying the first weights to generate the second weights.

The first weights may be modified by applying an exponential or power-law function to the first weights in order to generate the second weights. The exponential or power-law function may at least one of: reduce first ones of the first weights or increase second ones of the first weights. The first ones of the first weights may have smaller values than the second ones of the first weights. The exponential or power-law function may cause the second weights, compared to the first weights, to favor the chroma channels of one or more first ones of the image frames over the chroma channels of one or more second ones of the image frames during the blending of the image frames. The one or more first ones of the image frames may have a lower exposure than the one or more second ones of the image frames.

The first weights may be modified by determining error maps based on hues of pairs of the image frames and modifying the first weights based on the error maps. In each pair of the image frames, one of the image frames in the pair may have a lower exposure than another of the image frames in the pair.

The first weights and the second weights may be determined by determining the first weights using one or more first tuning parameters and separately determining the second weights using one or more second tuning parameters. The one or more second tuning parameters may be different than the one or more first tuning parameters.

The second weights may be determined using the one or more second tuning parameters by determining different initial weights for the image frames and combining the different initial weights for the image frames to generate the second weights. The different initial weights may be associated with different characteristics of the image frames.

The different initial weights for the image frames may include color saturation weights associated with the image frames. The color saturation weights may be based on amounts of color saturation within the image frames. The different initial weights for the image frames may be combined by at least one of: reducing ones of the initial weights that are associated with larger amounts of color saturation within the image frames or increasing others of the initial weights that are associated with smaller amounts of color saturation within the image frames.

The image frames may include at least one lower-exposure image frame and at least one higher-exposure image frame. The second weights may be determined in order to, during the blending of the image frames, at least one of: (i) increase contribution of the chroma channels of the at least one lower-exposure image frame or (ii) decrease contribution of the chroma channels of the at least one higher-exposure image frame.

Other technical features may be readily apparent to one skilled in the art from the following figures, descriptions, and claims.

Before undertaking the DETAILED DESCRIPTION below, it may be advantageous to set forth definitions of certain words and phrases used throughout this patent document. The terms “transmit,” “receive,” and “communicate,” as well as derivatives thereof, encompass both direct and indirect communication. The terms “include” and “comprise,” as well as derivatives thereof, mean inclusion without limitation. The term “or” is inclusive, meaning and/or. The phrase “associated with,” as well as derivatives thereof, means to include, be included within, interconnect with, contain, be contained within, connect to or with, couple to or with, be communicable with, cooperate with, interleave, juxtapose, be proximate to, be bound to or with, have, have a property of, have a relationship to or with, or the like.

Moreover, various functions described below can be implemented or supported by one or more computer programs, each of which is formed from computer readable program code and embodied in a computer readable medium. The terms “application” and “program” refer to one or more computer programs, software components, sets of instructions, procedures, functions, objects, classes, instances, related data, or a portion thereof adapted for implementation in a suitable computer readable program code. The phrase “computer readable program code” includes any type of computer code, including source code, object code, and executable code. The phrase “computer readable medium” includes any type of medium capable of being accessed by a computer, such as read only memory (ROM), random access memory (RAM), a hard disk drive, a compact disc (CD), a digital video disc (DVD), or any other type of memory. A “non-transitory” computer readable medium excludes wired, wireless, optical, or other communication links that transport transitory electrical or other signals. A non-transitory computer readable medium includes media where data can be permanently stored and media where data can be stored and later overwritten, such as a rewritable optical disc or an erasable memory device.

As used here, terms and phrases such as “have,” “may have,” “include,” or “may include” a feature (like a number, function, operation, or component such as a part) indicate the existence of the feature and do not exclude the existence of other features. Also, as used here, the phrases “A or B,” “at least one of A and/or B,” or “one or more of A and/or B” may include all possible combinations of A and B. For example, “A or B,” “at least one of A and B,” and “at least one of A or B” may indicate all of (1) including at least one A, (2) including at least one B, or (3) including at least one A and at least one B. Further, as used here, the terms “first” and “second” may modify various components regardless of importance and do not limit the components. These terms are only used to distinguish one component from another. For example, a first user device and a second user device may indicate different user devices from each other, regardless of the order or importance of the devices. A first component may be denoted a second component and vice versa without departing from the scope of this disclosure.

It will be understood that, when an element (such as a first element) is referred to as being (operatively or communicatively) “coupled with/to” or “connected with/to” another element (such as a second element), it can be coupled or connected with/to the other element directly or via a third element. In contrast, it will be understood that, when an element (such as a first element) is referred to as being “directly coupled with/to” or “directly connected with/to” another element (such as a second element), no other element (such as a third element) intervenes between the element and the other element.

As used here, the phrase “configured (or set) to” may be interchangeably used with the phrases “suitable for,” “having the capacity to,” “designed to,” “adapted to,” “made to,” or “capable of” depending on the circumstances. The phrase “configured (or set) to” does not essentially mean “specifically designed in hardware to.” Rather, the phrase “configured to” may mean that a device can perform an operation together with another device or parts. For example, the phrase “processor configured (or set) to perform A, B, and C” may mean a generic-purpose processor (such as a CPU or application processor) that may perform the operations by executing one or more software programs stored in a memory device or a dedicated processor (such as an embedded processor) for performing the operations.

The terms and phrases as used here are provided merely to describe some embodiments of this disclosure but not to limit the scope of other embodiments of this disclosure. It is to be understood that the singular forms “a,” “an,” and “the” include plural references unless the context clearly dictates otherwise. All terms and phrases, including technical and scientific terms and phrases, used here have the same meanings as commonly understood by one of ordinary skill in the art to which the embodiments of this disclosure belong. It will be further understood that terms and phrases, such as those defined in commonly-used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined here. In some cases, the terms and phrases defined here may be interpreted to exclude embodiments of this disclosure.

Examples of an “electronic device” according to embodiments of this disclosure may include at least one of a smartphone, a tablet personal computer (PC), a mobile phone, a video phone, an e-book reader, a desktop PC, a laptop computer, a netbook computer, a workstation, a personal digital assistant (PDA), a portable multimedia player (PMP), an MP3 player, a mobile medical device, a camera, or a wearable device (such as smart glasses, a head-mounted device (HMD), electronic clothes, an electronic bracelet, an electronic necklace, an electronic accessory, an electronic tattoo, a smart mirror, or a smart watch). Other examples of an electronic device include a smart home appliance. Examples of the smart home appliance may include at least one of a television, a digital video disc (DVD) player, an audio player, a refrigerator, an air conditioner, a cleaner, an oven, a microwave oven, a washer, a dryer, an air cleaner, a set-top box, a home automation control panel, a security control panel, a TV box (such as SAMSUNG HOMESYNC, APPLETV, or GOOGLE TV), a smart speaker or speaker with an integrated digital assistant (such as SAMSUNG GALAXY HOME, APPLE HOMEPOD, or AMAZON ECHO), a gaming console (such as an XBOX, PLAYSTATION, or NINTENDO), an electronic dictionary, an electronic key, a camcorder, or an electronic picture frame. Still other examples of an electronic device include at least one of various medical devices (such as diverse portable medical measuring devices (like a blood sugar measuring device, a heartbeat measuring device, or a body temperature measuring device), a magnetic resource angiography (MRA) device, a magnetic resource imaging (MRI) device, a computed tomography (CT) device, an imaging device, or an ultrasonic device), a navigation device, a global positioning system (GPS) receiver, an event data recorder (EDR), a flight data recorder (FDR), an automotive infotainment device, a sailing electronic device (such as a sailing navigation device or a gyro compass), avionics, security devices, vehicular head units, industrial or home robots, automatic teller machines (ATMs), point of sales (POS) devices, or Internet of Things (IoT) devices (such as a bulb, various sensors, electric or gas meter, sprinkler, fire alarm, thermostat, street light, toaster, fitness equipment, hot water tank, heater, or boiler). Other examples of an electronic device include at least one part of a piece of furniture or building/structure, an electronic board, an electronic signature receiving device, a projector, or various measurement devices (such as devices for measuring water, electricity, gas, or electromagnetic waves). Note that, according to various embodiments of this disclosure, an electronic device may be one or a combination of the above-listed devices. According to some embodiments of this disclosure, the electronic device may be a flexible electronic device. The electronic device disclosed here is not limited to the above-listed devices and may include any other electronic devices now known or later developed.

In the following description, electronic devices are described with reference to the accompanying drawings, according to various embodiments of this disclosure. As used here, the term “user” may denote a human or another device (such as an artificial intelligent electronic device) using the electronic device.

Definitions for other certain words and phrases may be provided throughout this patent document. Those of ordinary skill in the art should understand that in many if not most instances, such definitions apply to prior as well as future uses of such defined words and phrases.

None of the description in this application should be read as implying that any particular element, step, or function is an essential element that must be included in the claim scope. The scope of patented subject matter is defined only by the claims. Moreover, none of the claims is intended to invoke 35 U.S.C. § 112 (f) unless the exact words “means for” are followed by a participle. Use of any other term, including without limitation “mechanism,” “module,” “device,” “unit,” “component,” “element,” “member,” “apparatus,” “machine,” “system,” “processor,” or “controller,” within a claim is understood by the Applicant to refer to structures known to those skilled in the relevant art and is not intended to invoke 35 U.S.C. § 112 (f).

1 10 FIGS.through , discussed below, and the various embodiments of this disclosure are described with reference to the accompanying drawings. However, it should be appreciated that this disclosure is not limited to these embodiments, and all changes and/or equivalents or replacements thereto also belong to the scope of this disclosure. The same or similar reference denotations may be used to refer to the same or similar elements throughout the specification and the drawings.

As noted above, many mobile electronic devices, such as smartphones and tablet computers, include cameras that can be used to capture still and video images. These types of devices often include image processing pipelines that perform a number of operations in sequence to generate images of scenes. For example, an image processing pipeline may perform operations that combine multiple image frames of a scene into a combined image and process the combined image to produce a final image of the scene.

One common feature of an image processing pipeline is a tone mapping operation, which can be used to adjust the colors in images produced by the image processing pipeline. Tone mapping can be useful or important in various applications, such as when image processing can result in the creation of unnatural tone within images. One example technique for tone mapping is synthetic exposure fusion-based tone mapping, which generates multiple image frames having different (typically synthetic) exposures and combines the image frames based on a blending map. The blending map can contain blending weights or other values that identify how pixels in different image frames are weighted during the combination of the image frames. One goal of synthetic exposure fusion-based tone mapping can be to combine multiple lower dynamic range image frames into a single image with improved colors.

Unfortunately, existing synthetic exposure fusion-based tone mapping techniques can produce image artifacts in generated images. For example, a wall or other object having substantially-uniform color may appear to have different color tones within a final image of a scene that includes the object. Among other reasons, existing synthetic exposure fusion-based tone mapping techniques use a single set of weights for all data channels of the image frames, which can cause color distortions even with a perfect identification of a blending map. Also, existing techniques for determining blending maps often do not take color distortions into account when determining the blending maps, which can lead to color reproduction issues when processing certain image frames (such as image frames containing colors that are highly saturated in one or more data channels).

Because of these types of color distortion issues, many synthetic exposure fusion-based tone mapping techniques are luma-based, meaning these techniques operate on the luma channels of image frames and not on chroma channels of the image frames. However, these approaches still have issues with color gamut clipping and distortion due to (i) image brightness being too high for a synthetic exposure and/or (ii) color saturation being too high. Simply reducing the contribution of higher-exposure image frames (which tend to suffer from over-saturation due to brightness) may not be effective since there is often a desire to have higher brightness in final output images. Also, reducing the contribution of higher-exposure image frames may be difficult since saturated colors often do not have large luma values, and current synthetic exposure fusion-based tone mapping techniques often favor higher exposures when luma pixels are not saturated.

This disclosure provides various techniques supporting color distortion-aware exposure fusion. As described in more detail below, multiple image frames of a scene can be obtained, and each image frame can include a luma channel and chroma channels. For example, the image frames may be obtained by generating the image frames based on an image produced in an image processing pipeline. In some cases, the image frames can have different (possibly synthetic) exposures. The image frames can be blended to generate a blended image. To blend the image frames, first weights to be applied when blending the luma channels of the image frames can be determined, and separate second weights to be applied when blending the chroma channels of the image frames can be determined. In some cases, the first weights may be based on the image frames, and the first weights may be modified (such as by using an exponential or power-law function and/or hue-based error maps) in order to generate the second weights. A fused luma image can be generated using the luma channels of the image frames and the first weights, such as by blending the luma channels of the image frames based on the first weights. At least one fused chroma image can be separately generated using the chroma channels of the image frames and the second weights, such as by blending the chroma channels of the image frames based on the second weights. The fused luma image and the at least one fused chroma image can be combined to generate a final image of the scene, and one or more post-processing operations may be performed if needed or desired.

In this way, the described techniques can be used to perform color distortion-aware exposure fusion, meaning the described techniques support synthetic exposure fusion-based tone mapping that reduces or avoids problems associated with color distortions. Among other reasons, this can be achieved by supporting different weights for luma and chroma channels of image frames. Also, for chroma-based weighting, separate weighting schemes may be provided for combining chroma channels of different synthetic exposures, such as when higher weights are given to lower-exposure image frames (since there is less color distortion in those image frames). In addition, saturated color detection can be performed, where pixels with saturated colors can be identified and where contributions from higher-exposure image frames for those pixels can be reduced (such as by using lower weights). Because of this, images produced using the described techniques can have significantly-improved quality and significantly-reduced color distortions.

Note that while various embodiments of this disclosure are described in the context of use with certain consumer electronic devices (such as smartphones or tablet computers), this is merely one example. It will be understood that the principles of this disclosure may be implemented in any number of other suitable contexts and may use any suitable device or devices. In general, this disclosure is not limited to use with any specific type(s) or number(s) of device(s).

1 FIG. 1 FIG. 100 100 100 illustrates an example network configurationincluding an electronic device in accordance with this disclosure. The embodiment of the network configurationshown inis for illustration only. Other embodiments of the network configurationcould be used without departing from the scope of this disclosure.

101 100 101 110 120 130 150 160 170 180 101 110 120 180 According to embodiments of this disclosure, an electronic deviceis included in the network configuration. The electronic devicecan include at least one of a bus, a processor, a memory, an input/output (I/O) interface, a display, a communication interface, and a sensor. In some embodiments, the electronic devicemay exclude at least one of these components or may add at least one other component. The busincludes a circuit for connecting the components-with one another and for transferring communications (such as control messages and/or data) between the components.

120 120 120 101 120 The processorincludes one or more processing devices, such as one or more microprocessors, microcontrollers, digital signal processors (DSPs), application specific integrated circuits (ASICs), or field programmable gate arrays (FPGAs). In some embodiments, the processorincludes one or more of a central processing unit (CPU), an application processor (AP), a communication processor (CP), a graphics processor unit (GPU), or a neural processing unit (NPU). The processoris able to perform control on at least one of the other components of the electronic deviceand/or perform an operation or data processing relating to communication or other functions. As described below, the processormay perform one or more functions related to color distortion-aware exposure fusion.

130 130 101 130 140 140 141 143 145 147 141 143 145 The memorycan include a volatile and/or non-volatile memory. For example, the memorycan store commands or data related to at least one other component of the electronic device. According to embodiments of this disclosure, the memorycan store software and/or a program. The programincludes, for example, a kernel, middleware, an application programming interface (API), and/or an application program (or “application”). At least a portion of the kernel, middleware, or APImay be denoted an operating system (OS).

141 110 120 130 143 145 147 141 143 145 147 101 147 143 145 147 141 147 143 147 101 110 120 130 147 145 147 141 143 145 The kernelcan control or manage system resources (such as the bus, processor, or memory) used to perform operations or functions implemented in other programs (such as the middleware, API, or application). The kernelprovides an interface that allows the middleware, the API, or the applicationto access the individual components of the electronic deviceto control or manage the system resources. The applicationmay include one or more applications that, among other things, perform color distortion-aware exposure fusion. These functions can be performed by a single application or by multiple applications that each carries out one or more of these functions. The middlewarecan function as a relay to allow the APIor the applicationto communicate data with the kernel, for instance. A plurality of applicationscan be provided. The middlewareis able to control work requests received from the applications, such as by allocating the priority of using the system resources of the electronic device(like the bus, the processor, or the memory) to at least one of the plurality of applications. The APIis an interface allowing the applicationto control functions provided from the kernelor the middleware. For example, the APIincludes at least one interface or function (such as a command) for filing control, window control, image processing, or text control.

150 101 150 101 The I/O interfaceserves as an interface that can, for example, transfer commands or data input from a user or other external devices to other component(s) of the electronic device. The I/O interfacecan also output commands or data received from other component(s) of the electronic deviceto the user or the other external device.

160 160 160 160 The displayincludes, for example, a liquid crystal display (LCD), a light emitting diode (LED) display, an organic light emitting diode (OLED) display, a quantum-dot light emitting diode (QLED) display, a microelectromechanical systems (MEMS) display, or an electronic paper display. The displaycan also be a depth-aware display, such as a multi-focal display. The displayis able to display, for example, various contents (such as text, images, videos, icons, or symbols) to the user. The displaycan include a touchscreen and may receive, for example, a touch, gesture, proximity, or hovering input using an electronic pen or a body portion of the user.

170 101 102 104 106 170 162 164 170 The communication interface, for example, is able to set up communication between the electronic deviceand an external electronic device (such as a first electronic device, a second electronic device, or a server). For example, the communication interfacecan be connected with a networkorthrough wireless or wired communication to communicate with the external electronic device. The communication interfacecan be a wired or wireless transceiver or any other component for transmitting and receiving signals.

162 164 The wireless communication is able to use at least one of, for example, WiFi, long term evolution (LTE), long term evolution-advanced (LTE-A), 5th generation wireless system (5G), millimeter-wave or 60 GHz wireless communication, Wireless USB, code division multiple access (CDMA), wideband code division multiple access (WCDMA), universal mobile telecommunication system (UMTS), wireless broadband (WiBro), or global system for mobile communication (GSM), as a communication protocol. The wired connection can include, for example, at least one of a universal serial bus (USB), high definition multimedia interface (HDMI), recommended standard 232 (RS-232), or plain old telephone service (POTS). The networkorincludes at least one communication network, such as a computer network (like a local area network (LAN) or wide area network (WAN)), Internet, or a telephone network.

101 180 101 180 180 180 180 180 101 The electronic devicefurther includes one or more sensorsthat can meter a physical quantity or detect an activation state of the electronic deviceand convert metered or detected information into an electrical signal. For example, the sensor(s)can include one or more cameras or other imaging sensors, which may be used to capture images of scenes. The sensor(s)can also include one or more buttons for touch input, one or more microphones, a depth sensor, a gesture sensor, a gyroscope or gyro sensor, an air pressure sensor, a magnetic sensor or magnetometer, an acceleration sensor or accelerometer, a grip sensor, a proximity sensor, a color sensor (such as a red green blue (RGB) sensor), a bio-physical sensor, a temperature sensor, a humidity sensor, an illumination sensor, an ultraviolet (UV) sensor, an electromyography (EMG) sensor, an electroencephalogram (EEG) sensor, an electrocardiogram (ECG) sensor, an infrared (IR) sensor, an ultrasound sensor, an iris sensor, or a fingerprint sensor. Moreover, the sensor(s)can include one or more position sensors, such as an inertial measurement unit that can include one or more accelerometers, gyroscopes, and other components. In addition, the sensor(s)can include a control circuit for controlling at least one of the sensors included here. Any of these sensor(s)can be located within the electronic device.

101 101 102 104 101 102 101 102 170 101 102 102 In some embodiments, the electronic devicecan be a wearable device or an electronic device-mountable wearable device (such as an HMD). For example, the electronic devicemay represent an XR wearable device, such as a headset or smart eyeglasses. In other embodiments, the first external electronic deviceor the second external electronic devicecan be a wearable device or an electronic device-mountable wearable device (such as an HMD). In those other embodiments, when the electronic deviceis mounted in the electronic device(such as the HMD), the electronic devicecan communicate with the electronic devicethrough the communication interface. The electronic devicecan be directly connected with the electronic deviceto communicate with the electronic devicewithout involving with a separate network.

102 104 106 101 106 101 102 104 106 101 101 102 104 106 102 104 106 101 101 101 170 104 106 162 164 101 1 FIG. The first and second external electronic devicesandand the servereach can be a device of the same or a different type from the electronic device. According to certain embodiments of this disclosure, the serverincludes a group of one or more servers. Also, according to certain embodiments of this disclosure, all or some of the operations executed on the electronic devicecan be executed on another or multiple other electronic devices (such as the electronic devicesandor server). Further, according to certain embodiments of this disclosure, when the electronic deviceshould perform some function or service automatically or at a request, the electronic device, instead of executing the function or service on its own or additionally, can request another device (such as electronic devicesandor server) to perform at least some functions associated therewith. The other electronic device (such as electronic devicesandor server) is able to execute the requested functions or additional functions and transfer a result of the execution to the electronic device. The electronic devicecan provide a requested function or service by processing the received result as it is or additionally. To that end, a cloud computing, distributed computing, or client-server computing technique may be used, for example. Whileshows that the electronic deviceincludes the communication interfaceto communicate with the external electronic deviceor servervia the networkor, the electronic devicemay be independently operated without a separate communication function according to some embodiments of this disclosure.

106 101 106 101 101 106 120 101 106 The servercan include the same or similar components as the electronic device(or a suitable subset thereof). The servercan support to drive the electronic deviceby performing at least one of operations (or functions) implemented on the electronic device. For example, the servercan include a processing module or processor that may support the processorimplemented in the electronic device. As described below, the servermay perform one or more functions related to color distortion-aware exposure fusion.

1 FIG. 1 FIG. 1 FIG. 1 FIG. 100 101 100 Althoughillustrates one example of a network configurationincluding an electronic device, various changes may be made to. For example, the network configurationcould include any number of each component in any suitable arrangement. In general, computing and communication systems come in a wide variety of configurations, anddoes not limit the scope of this disclosure to any particular configuration. Also, whileillustrates one operational environment in which various features disclosed in this patent document can be used, these features could be used in any other suitable system.

2 2 FIGS.A andB 2 FIG.A 2 FIG.B 2 2 FIGS.A andB 1 FIG. 200 200 200 200 101 100 200 106 illustrate an example pipelinesupporting color distortion-aware exposure fusion in accordance with this disclosure. More specifically,illustrates the example pipeline, andillustrates one operation within the pipelinein greater detail. For ease of explanation, the pipelineshown inis described as being implemented using the electronic devicein the network configurationshown in. However, the pipelinemay be implemented using any other suitable device(s) (such as the server) and in any other suitable system(s).

2 FIG.A 200 202 202 202 180 101 202 202 202 202 As shown in, the image processing pipelinegenerally receives and processes input image frames. Each input image framemay be obtained from any suitable source, such as when the input image framesare produced by at least one camera or other imaging sensorof the electronic deviceduring an image capture operation. In some embodiments, the input image framesmay represent raw image frames. Raw image frames typically refer to image frames that have undergone little if any processing after being captured. The availability of raw image frames can be useful in a number of circumstances since the raw image frames can be subsequently processed to achieve the creation of desired effects in output images. In many cases, for example, the input image framescan have a wider dynamic range or a wider color gamut that is narrowed during image processing operations in order to produce still or video images suitable for display or other use. Each input image framecan have any suitable format, such as a Bayer or other raw image format, a red-green-blue (RGB) image format, or a luma-chroma (YUV) image format. Each input image framecan also have any suitable resolution, such as up to fifty megapixels or more.

202 101 202 180 202 202 202 In some embodiments, the input image framesmay include two or more image frames captured using different capture conditions. The capture conditions can represent any suitable settings of the electronic deviceor other device used to capture the input image framesor any suitable contents of scenes being imaged. For example, the capture conditions may represent different exposure settings of the imaging sensor(s)used to capture the input image frames, such as different exposure times or ISO settings. In multi-frame processing pipelines, for instance, different input image framescan be captured using different exposure settings so that portions of different input image framescan be combined to produce a high dynamic range (HDR) output image or other blended image.

202 200 180 202 In some embodiments, the input image framesmay undergo one or more pre-processing operations prior to further processing in the pipeline. Any suitable pre-processing operation(s) may be performed here. Examples of pre-processing operations may include bad pixel correction (identifying and replacing bad pixel data, such as via interpolation of neighboring good pixel data), lens shading correction (compensating for peripheral shading created by one or more lenses used in or with one or more imaging sensors), and/or white balance adjustment (modifying the white balance of one or more input image frames). Note, however, that this disclosure is not limited to any particular image pre-processing technique(s).

202 204 202 202 202 202 202 202 101 202 202 The input image frames(or possibly preprocessed versions thereof) are provided to an image registration and blending operation, which generally operates to (i) modify one or more of the input image framesin order to generate aligned versions of the input image framesand (ii) blend or otherwise combine the aligned versions of the input image frames. For example, the input image framesmay undergo registration so that common features in different input image framesare at the same or substantially the same locations in the aligned versions of the input image frames. In some embodiments, registration may be performed by selecting a reference image frame and modifying one or more non-reference image frames so as to be aligned with the reference image frame, such as by generating a warp or alignment map for each non-reference image frame, where each warp or alignment map includes or is based on one or more motion vectors that identify how the position(s) of one or more specific features in the associated non-reference image frame should be altered in order to be in the position(s) of the same feature(s) in the reference image frame. Registration may be needed in order to compensate for misalignment caused by the electronic devicemoving or rotating in between image captures, which causes objects in the input image framesto move or rotate slightly (as is common with handheld devices). Registration, which is also sometimes referred to as image alignment, may be performed using any suitable technique(s). In some embodiments, the input image framescan be aligned both geometrically and photometrically. In particular embodiments, the registration can be performed using global Oriented FAST and Rotated BRIEF (ORB) features and local features from a block search to identify how to align the image frames. Note, however, that this disclosure is not limited to any particular technique(s) for aligning image frames.

202 202 202 The blending generally operates to combine image data contained in the aligned input image framesin order to generate a blended image. For instance, blending may occur by processing the aligned input image framesin order to modify portions of the selected reference frame using image data from one or more non-reference frames. As a particular example, blending may take the reference frame and replace one or more portions of the reference frame containing motion with one or more corresponding portions of one or more lower-exposure image frames, thereby producing the blended image. In some cases, the blending may involve a weighted blending operation that combines the pixel values contained in the aligned image framesbased on weights. Note, however, that this disclosure is not limited to any particular technique(s) for combining image frames.

206 206 206 The blended image is provided to a demosaicing operation, which generally operates to convert image data produced using a Bayer filter array or other color filter array into reconstructed red-green-blue (RGB) data or other image data in order to generate a demosaiced image. For example, the demosaicing operationcan perform various interpolations to fill in missing information, such as by estimating other colors' image data for each pixel. When using a Bayer filter array or some other types of color filter arrays, approximately twice as many pixels may capture image data using green filters compared to pixels that capture image data using red or blue filters. This can introduce non-uniformities into the captured image data, such as when the red and blue image data each have a lower signal-to-noise ratio (SNR) and a lower sampling rate compared to the green image data. Among other things, the green image data can capture high-frequency image content more effectively than the red and blue image data. The demosaicing operationcan take information captured by at least one highly-sampled channel (such as the green channel and/or the white channel) and use that information to correct limitations of lower-sampled channels (such as the red and blue channels), which can help to reintroduce high-frequency image content into the red and blue image data. Note, however, that this disclosure is not limited to any particular technique(s) for demosaicing images.

208 208 208 208 The demosaiced image is provided to a denoising operation, which generally operates to process the demosaiced image and remove noise from the demosaiced image in order to generate a filtered image, such as an HDR image. For example, the denoising operationmay be used to remove sampling, interpolation, and aliasing artifacts and noise in subsampled image channels (such as the red and blue channels) of the demosaiced image using information from at least one higher-sampled channel (such as the green channel and/or the white channel) of the demosaiced image. The denoising operationmay also or alternatively be used to filter the image data of the demosaiced image in order to remove noise from object edges, which can help to provide cleaner edges to objects captured in the demosaiced image. The denoising operationmay use any suitable technique(s) for filtering image data, such as spatial noise filtering. Note, however, that this disclosure is not limited to any particular technique(s) for filtering image data.

210 202 210 210 212 212 The filtered image is provided to a tone mapping operation, which generally operates to adjust colors in the filtered image. As noted above, this can be useful or important in various applications, such as when generating HDR images. For instance, since generating an HDR image often involves capturing multiple image framesof a scene using different exposures and combining the captured image frames to produce the HDR image, this type of processing can often result in the creation of unnatural tone within the HDR image. The tone mapping operationcan therefore adjust the colors contained in the filtered image, such as by performing dynamic range compression, in order to provide more natural colors or other results. The output of the tone mapping operationcan represent an output image, which may represent a final image of the scene. Note, however, that the output imagemay undergo one or more post-processing operations to produce a final image of the scene.

2 FIG.B 2 FIG.B 210 200 210 250 204 208 200 250 illustrates an example implementation of the tone mapping operationof the pipeline. As shown in, the tone mapping operationobtains a multi-exposed image, such as an HDR image produced by the operations-in the pipeline. In some cases, the multi-exposed imagemay represent a linear HDR image.

210 252 250 254 254 250 254 250 252 254 254 254 252 252 252 254 2 FIG.B The tone mapping operationperforms one or more synthetic exposure generation operationsusing the multi-exposed imageto generate multiple synthetic-exposure image frames. Each synthetic-exposure image framerepresents a synthetic or machine-generated version of the multi-exposed imageat a specified exposure. Different synthetic-exposure image framesrepresent synthetic or machine-generated versions of the multi-exposed imageat different exposures. For example, the synthetic exposure generation operation(s)may operate to produce one or more lower-exposure image frames, one or more medium-exposure image frames, and one or more higher-exposure image frames. Note here that “lower,” “medium,” and “higher” are used relative to each other and do not require any specific values of exposure settings. Lower-exposure image frames are often associated with shorter exposure times and/or faster shutter speeds, medium-exposure image frames are often associated with intermediate exposure times and/or intermediate shutter speeds, and higher-exposure image frames are often associated with longer exposure times and/or slower shutter speeds. Smaller-exposure image frames can often capture details in brighter portions of a scene that would appear “blown out” at higher exposure values (where “blown out” encompasses conditions where light provided by a portion of a scene exceeds an upper limit of a dynamic range of an imaging sensor, resulting in significant or total loss of color saturation). Each synthetic exposure generation operationincludes any suitable logic for generating a synthetic image frame at a specified exposure. While multiple instances of the synthetic exposure generation operationare shown in, the same operationmay be used repeatedly (such as serially) to generate the synthetic-exposure image frames.

256 254 212 256 256 254 212 256 254 212 254 254 254 A fusion operationgenerally operates to fuse or otherwise combine the synthetic-exposure image framesin order to generate an output image. Example implementations of the fusion operationare described below. In some embodiments, the fusion operationmay perform weighted blending of pixel values in the synthetic-exposure image framesin order to generate pixel values in the output image. The fusion operationcan also perform weighted blending of luma and chroma channels of the synthetic-exposure image framesseparately in order to generate a fused luma image and at least one fused chroma image, which can be combined to produce the output image. As described below, the luma and chroma channels of the synthetic-exposure image framescan be weighted differently during the fusion. Also, for chroma-based weighting, separate weighting schemes may be used to combine the chroma channels of the synthetic-exposure image frames, such as to give higher weights to synthetic-exposure image framesassociated with lower exposures (where there is less color distortion).

256 254 256 254 254 254 254 254 254 254 254 254 254 254 The weights used by the fusion operationto combine the luma and chroma channels of the synthetic-exposure image framescan be determined in any suitable manner. For example, in some embodiments, the fusion operationmay determine first weights used to combine the luma channels of the synthetic-exposure image frames, and at least some of the first weights can be modified in order to generate second weights used to combine the chroma channels of the synthetic-exposure image frames. In these embodiments, the first weights could be modified to favor synthetic-exposure image frameshaving lower exposures over synthetic-exposure image frameshaving higher exposures under certain circumstances. In particular embodiments, the first weights may be modified by applying an exponential or power-law function to the first weights in order to generate the second weights, where (i) the exponential or power-law function reduces first ones of the first weights and/or increases second ones of the first weights and (ii) the first ones of the first weights have smaller values than the second ones of the first weights. Here, the exponential or power-law function causes the second weights, compared to the first weights, to favor the chroma channels of one or more first ones of the synthetic-exposure image framesover the chroma channels of one or more second ones of the synthetic-exposure image framesduring the blending of the synthetic-exposure image frames, where the one or more first ones of the synthetic-exposure image frameshave a lower exposure than the one or more second ones of the synthetic-exposure image frames. As another particular example, the first weights may be modified by determining error maps based on hues of pairs of the synthetic-exposure image framesand modifying the first weights based on the error maps, where each pair of the synthetic-exposure image framesincludes an image frame having a lower exposure than the other image frame.

256 254 254 254 254 254 254 254 254 254 In other embodiments, the fusion operationmay determine first weights used to combine the luma channels of the synthetic-exposure image framesand separately determine second weights used to combine the chroma channels of the synthetic-exposure image frames. In these embodiments, the first and second weights could be generated using different tuning parameters. For example, initial weights may be determined for the synthetic-exposure image frames, such as based on different characteristics of the synthetic-exposure image frames. The initial weights for the synthetic-exposure image framescan be combined to generate the first weights used to blend the luma channels of the synthetic-exposure image frames. Similar operations may occur to generate the second weights used to blend the chroma channels of the synthetic-exposure image frames. Again, in some cases, this can be done to favor synthetic-exposure image frameshaving lower exposures over synthetic-exposure image frameshaving higher exposures.

256 254 254 254 254 254 254 254 254 The fusion operationcan also identify pixels of the synthetic-exposure image framesassociated with or containing saturated colors. For example, color saturation weights associated with the synthetic-exposure image framescan be identified, where the color saturation weights are based on amounts of color saturation within the synthetic-exposure image frames. The weights used to combine the luma channels and/or the chroma channels of the synthetic-exposure image framescan be modified based on the color saturation weights, such as by reducing ones of the initial weights that are associated with larger amounts of color saturation within the synthetic-exposure image framesand/or increasing others of the initial weights that are associated with smaller amounts of color saturation within the synthetic-exposure image frames. In some cases, this allows contributions from saturated pixels in higher-exposure synthetic-exposure image framesto be reduced, thereby favoring pixels in lower-exposure synthetic-exposure image frames, when color saturation is detected.

2 2 FIGS.A andB 2 2 FIGS.A andB 2 2 FIGS.A andB 2 2 FIGS.A andB 200 200 Althoughillustrate one example of a pipelinesupporting color distortion-aware exposure fusion, various changes may be made to. For example, various components and operations inmay be combined, further subdivided, replicated, rearranged, or omitted according to particular needs. Also, various additional components and operations may be used in. In addition, the specific image processing pipelinedescribed above is for illustration and explanation only. Various image processing pipelines have been developed, and additional image processing pipelines are sure to be developed in the future. This disclosure is not limited to any specific implementation of an image processing pipeline or even to use within an image processing pipeline. In general, the techniques for color distortion-aware exposure fusion described in this patent document may be used in any other image processing pipeline or other architecture.

3 FIG. 3 FIG. 1 FIG. 2 2 FIGS.A andB 3 FIG. 300 300 101 100 210 256 300 106 illustrates a first example architecturefor color distortion-aware exposure fusion in accordance with this disclosure. For ease of explanation, the architectureshown inis described as being used by the electronic devicein the network configurationshown inin order to implement at least part of the tone mapping operation(such as the fusion operation) shown in. However, the architecturemay be used with any other suitable device(s) (such as the server) or pipeline(s) and in any other suitable system(s). In, note that thicker lines are used to denote multiple images or image frames and thinner lines are used to denote single images or image frames.

3 FIG. 302 302 302 254 302 302 As shown in, multiple multi-exposed image framesare obtained, where different ones of the multi-exposed image framesare associated with different exposures. The multi-exposed image framesmay, for example, represent the synthetic-exposure image frames. In some embodiments, each multi-exposed image frameincludes a luma channel and multiple chroma channels, meaning there is pixel data in a luma channel and pixel data in different chroma channels. In some cases, the pixel data of each multi-exposed image framemay be denoted as (Y,Cb,Cr) or YCbCr, where Y represents the luma (luminance) channel data and Cb and Cr represent the chroma (blue-difference chrominance and red-difference chrominance) channel data.

302 304 302 304 302 302 304 302 The multi-exposed image framesare provided to a blending map generation operation, which generally operates to produce one or more initial blending maps for the multi-exposed image frames. In some embodiments, the blending map generation operationcan produce an initial blending map for each multi-exposed image frame, where each initial blending map identifies initial contributions of pixel values in the associated multi-exposed image frameto a final output image. The blending map generation operationcan use any suitable technique(s) to generate initial blending maps for multi-exposed image frames.

th th th th th 302 302 302 302 302 302 304 In some embodiments, consider the imulti-exposed image framein a set of multi-exposed image frames, where the imulti-exposed image frameis associated with a lower exposure (such as a shorter exposure time). All other things being equal, image details in any darker region of the imulti-exposed image frameare more likely to be poor or nonexistent due to the shorter exposure time, while image details in any brighter regions of the imulti-exposed image frameare more likely to exhibit good color saturation and not be blown out or over-saturated. Thus, within an initial blending map for the imulti-exposed image frame, the blending map generation operationcould assign higher weights to brighter regions of the image frame than to darker regions of the image frame. In this way, the contribution of the best-exposed portions of the image frame to the final output image could be greater than the contribution of the poorly-exposed portions of the image frame.

304 302 304 302 th th th i In particular embodiments, the blending map generation operationmay operate as follows. For each imulti-exposed image frame, the initial blending map generated by the blending map generation operationfor that image framemay include or be based on a composite of at least (i) a first map of values of a contrast or saliency metric C and (ii) a second map of values of a color saturation metric S. In this example, the values of the contrast or saliency metric C can correspond to the extent to which a given region of the 7th image frame exhibits sufficient contrast from which edge details can be perceived. Also, in this example, the values of the color saturation metric S can correspond to color saturation, such as whether colors in a particular region of the iimage frame appear deep or “blown out” and almost white. As a particular example, a first map Cof the saliency metric C for the iimage frame may be obtained as follows.

i th th th Here, Yrepresents the luma channel of the iimage frame, L represents a Laplacian operator, and G represents a Gaussian filter. As another particular example, where image data for the iimage frame is provided through the channels of the YCbCr color space, a second map Si of the color saturation metric S for the iimage frame may be obtained as follows.

i i i th th th Here, Cbrepresents the Cb chroma channel of the iimage frame, and Crrepresents the Cr chroma channel of the iimage frame. In some cases, an initial blending map Pfor the iimage frame may be generated by combining and normalizing the first and second maps, which could be expressed as follows.

i i i 302 Here, {tilde over (P)}represents the combined maps, Prepresents the combined maps after normalization, and N represents the number of multi-exposed image frames. These operations may be repeated for each image frame so that an initial blending map P is generated for each image frame. In this way, Prepresents channel-agnostic initial blending weights, which can be refined to properly register with object boundaries within a scene.

302 302 302 302 302 302 302 302 302 306 302 302 308 302 302 302 302 306 302 308 302 a a a a a a a a a. Luma channelsof the multi-exposed image framescan be processed using a set of operations to produce a fused luma image based on the multi-exposed image framesand the blending maps. For example, each of the luma channelsof the multi-exposed image framescan be decomposed into a base layer and a detail layer. In some embodiments, image data of the luma channelof each image framecan be represented as a superposition or sum of a base layer and a detail layer. A “base layer” refers to an image layer that includes large-scale (such as lower frequency) variations in values of a channel of image data, and a “detail layer” refers to an image layer that includes small-scale (such as higher frequency relative to the base layer) variations in values of the channel of image data. In this example, the luma channelsof the multi-exposed image framescan be processed using an average filtering operation, which can filter pixel values of the luma channelof each multi-exposed image frame. A difference operationcalculates differences between the luma channelsof the multi-exposed image framesand the associated average/filtered luma channelsof the multi-exposed image frames. Here, the outputs of the average filtering operationrepresent the base layers of the luma channels, and the outputs of the difference operationrepresent the detail layers of the luma channels

302 302 a In some embodiments, the image data forming the luma channelof the 7th image framemay be expressed as follows.

i i i i i Here, Yrepresents a sum of the detail layer YD; and the base layer YB. Thus, the detail layer YDcan be obtained by subtracting the base layer YBfrom the original channel image data Yas follows.

th th th 302 a i In particular embodiments, the base layer for the iimage frame can be obtained by applying a box filter (such as boxfilt or imboxfilt in MATLAB) to the luma channelof the iimage frame. For example, the base layer YBof the Y channel of the iimage frame could be obtained as follows.

300 302 302 i i i a Here, θ is a parameter setting the size or kernel of the box filter. In some cases, the value of θ can be tuned to optimize the quality of images produced by the architecture. From this, the detail layer YDcan be obtained by subtracting the base layer YBfrom Y. The base and detail layers of the luma channelfor each of the multi-exposed image framescan be generated in this manner.

i 304 302 In many cases, the initial blending maps Pproduced by the blending map generation operationare noisy, and transitions in blending weights can be spatially inconsistent relative to the boundaries of objects in a scene captured in the multi-exposed image frames. In practical terms, if the final output image is formed by blending image frames based on the initial blending maps, the weighting values for a given image frame or image frame channel may not register precisely with object boundaries within the scene, producing regions of improper exposures around the boundaries between brighter and darker areas of the scene. Left uncorrected, these registration errors can appear in the final output image as (i) light halos in which high weighting given to a higher-exposure image frame to bring out detail in a darker region spills over to a brighter region and/or (ii) dark halos in which high weighting given to a lower-exposure image frame to preserve detail in a brighter region spills over to a darker region.

304 310 304 312 304 310 314 314 302 302 314 302 312 316 316 302 302 316 302 318 302 In order to help compensate for these registration errors, one or more filters, such as one or more guided filters, can be applied to the blending maps from the blending map generation operation. In this example, a guided filtering operationcan apply a larger kernel to the blending maps from the blending map generation operation, and a guided filtering operationcan apply a smaller kernel to the blending maps from the blending map generation operation. The guided filtering operationprocesses the blending maps using the larger kernel to produce first filtered blending maps, which are provided to a weighted averaging operation. The weighted averaging operationapplies the first filtered blending maps to the base layers of the multi-exposed image framesin order to fuse the base layers of the multi-exposed image frames, and the weighted averaging operationoutputs a fused base layer for the multi-exposed image frames. Similarly, the guided filtering operationprocesses the blending maps using the smaller kernel to produce second filtered blending maps, which are provided to a weighted averaging operation. The weighted averaging operationapplies the second filtered blending maps to the detail layers of the multi-exposed image framesin order to fuse the detail layers of the multi-exposed image frames, and the weighted averaging operationoutputs a fused detail layer for the multi-exposed image frames. A combination operationcombines the fused base layer and the fused detail layer to produce a fused luma image for the multi-exposed image frames.

310 312 302 302 304 310 302 302 302 302 310 312 302 302 302 302 312 i i i i i a a a a The guided filtering operationsandhere can perform guided filtering of the blending maps with guidance from the multi-exposed image framesso that the filtered blending maps preserve edge and gradient information from the multi-exposed image frames. In some embodiments, for each of the blending maps Pgenerated by the blending map generation operation, alignment of blending weights relative to object boundaries can be enhanced by applying one or more guided filters operating as one or more edge-preserving filters, where each guided filter has a specified kernel size governing the size of the image frame area processed by the guided filter. The guided filtering operationuses a larger kernel (such as 25×25) that is applied to each initial blending map Pto obtain, for the luma channelof each multi-exposed image frame, a blending weight map WB; for the base layer. The data Yof the luma channelof each multi-exposed image framecan be used as a guiding image for the guided filtering operation. The guided filtering operationuses a smaller kernel (such as 5×5) that is applied to each initial blending map Pto obtain, for the luma channelof each multi-exposed image frame, a blending weight map WDi for the detail layer. Again, the data Yof the luma channelof each multi-exposed image framecan be used as a guiding image for the guided filtering operation.

302 302 302 302 302 302 302 302 304 320 320 304 302 302 302 302 302 302 302 a b b a b b a In this example, the weights applied to the luma channelsof the multi-exposed image framecan be said to represent first weights. In a similar manner, chroma channelsof the multi-exposed image framescan be processed using a set of operations to produce at least one fused chroma image based on the multi-exposed image framesand the blending maps, where the weights applied to the chroma channelsof the multi-exposed image framecan be said to represent second weights. In this example, the second weights are generated in a similar manner as for the luma channels, but the blending maps produced by the blending map generation operationare first modified using a blending map modification operation. The blending map modification operationcan process the blending maps produced by the blending map generation operation, possibly along with the multi-exposed image framesor other information, in order to modify at least some of the weights in the blending maps to generate modified blending maps for the chroma channelsof the multi-exposed image frame. In this way, it is possible to adjust the colors within the multi-exposed image frameby controlling how the chroma channelsare blended without making identical adjustments to the brightness (luma channels) within the multi-exposed image frame.

302 302 322 324 302 302 320 326 328 330 326 332 328 334 302 322 334 306 318 322 334 322 334 336 302 338 212 b b The chroma channelsof the multi-exposed image frameare provided to an average filtering operationand a difference operation, which respectively generate base and detail layers of the chroma channelsof the multi-exposed image frame. The modified blending maps produced by the blending map modification operationare provided to a guided filtering operationthat uses a larger kernel and a guided filtering operationthat uses a smaller kernel. A weighted averaging operationfuses the base layers based on filtered blending maps generated by the guided filtering operation, and a weighted averaging operationfuses the detail layers based on filtered blending maps generated by the guided filtering operation. A combination operationcombines the fused base layer(s) and the fused detail layer(s) to produce at least one fused chroma image for the multi-exposed image frames. These operations-can be performed in the same or similar manner as the operations-described above. In some cases, the operations-can be used to form multiple fused chroma images, such as a fused chroma image for the Cb channel and a fused chroma image for the Cr channel. In other cases, the operations-can be used to form a single fused chroma image for both chroma channels. A combination operationcombines the fused luma image and the at least one fused chroma image for the multi-exposed image framesto produce an output image, which in some cases could represent the output image.

320 302 302 300 320 402 304 320 404 402 406 404 402 406 302 302 302 302 a b b b 4 5 FIGS.and 3 FIG. 4 FIG. There are various ways in which the blending map modification operationmay be implemented in order to modify the blending weights for the luma channelsfor use with the chroma channels.illustrate example techniques for modifying blending maps in the architectureofin accordance with this disclosure. As shown in, the blending map modification operationreceives original blending maps, which represent the blending maps produced by the blending map generation operation. The blending map modification operationapplies an exponential or power-law functionto the weights of the original blending mapsin order to produce modified blending maps. The exponential or power-law functionhere can reduce certain weights (such as smaller weights) and/or increase certain weights (such as larger weights) contained in the original blending maps. As a result, the weights in the modified blending mapscan favor the chroma channelsof certain multi-exposed image frames(such as lower-exposure image frames) over the chroma channelsof other multi-exposed image frames(such as higher-exposure image frames).

402 302 As a particular example of this, assume that the original blending mapsfor N multi-exposed image framesas denoted as follows.

302 302 404 402 406 Here, index 0 could represent the image framewith the lowest exposure, and index N−1 could represent the image framewith the highest exposure. The exponential or power-law functioncan modify the weights of the original blending mapsto produce the modified blending maps, which can be denoted as follows.

i th 406 404 406 Here, {tilde over (W)}represents the imodified blending map. In some cases, the exponential or power-law functioncan apply the following function to generate the modified blending maps.

i i j i i 0 1 N-1 i th 402 320 302 302 302 b b a. Here, prepresents an exponent applied to the ioriginal blending map(denoted W), and prepresents an exponent used in a summation to normalize the result obtained by applying the exponent p. The exponents pare tunable, and p<p< . . . <p. In some embodiments, typical values of pcould be between one and five. Raising weights related to larger exposures to an exponential power reduces the resulting modified weights heavily, such as for regions where the weights are already not high compared to other exposures. This helps to ensure that brighter regions (where the higher exposures have color distortions) do not contribute significantly to final blending weights for the chroma channels and instead favor lower exposures to obtain chroma channel weights. Note that by using the blending map modification operationto modify the weights associated with the chroma channels, it is possible to apply separate corrections to the chroma channelswithout making similar corrections to the luma channels

5 FIG. 320 502 304 320 302 504 506 508 510 504 506 504 506 508 510 512 514 302 302 516 502 302 506 514 518 516 502 518 504 302 504 506 302 506 As shown in, the blending map modification operationagain receives original blending maps, which represent the blending maps produced by the blending map generation operation. The blending map modification operationalso receives pairs of the multi-exposed image frames, namely a lower-exposure image frameand a higher-exposure image frame. Hue calculation functionsandgenerally operate to determine hue maps for the image framesand. In some cases, each hue map could identify a hue of various pixels within the corresponding image frameor. Each hue calculation functionandcan use any suitable technique(s) to identify hue in an image frame, such as by performing hue/saturation/value (HSV) color-space transformation or by calculating a hue angle in a YCbCr transformation. The hue maps are provided to an error map generation function, which generally operates to compare the hue maps (such as by calculating pixel-wise differences between the hue maps) and generate an error mapidentifying the differences between the hue maps for that pair of multi-exposed image frames. These differences are indicative of color distortions in at least one of the two image framesin the pair. A weight modification functionmodifies the original blending mapcorresponding to one of the two image frames(such as the higher-exposure image frame) using the error mapto generate a modified blending map. For example, the weight modification functioncan modify weights from the original blending mapso that weights in the modified blending mapfavor the lower-exposure image frameduring blended. Note that this can be done for each of the image frames, such as when a lowest-exposure image frameis compared to each higher-exposure image framein the set of image framesin order to modify the blending map for each higher-exposure image frame.

518 As a particular example of this, the modified blending mapscan be calculated as follows.

514 i Here, GF represents a guided filtering function, and et represents the hue error in the error map. In some cases, the hue error ecan be defined as follows.

1 2 i o 1 504 506 Here, Tand Trepresent tunable thresholds, Yrepresents luma channel data, Hrepresents the hue map for the lower-exposure image frame, and Hrepresents the hue map for the higher-exposure image frame.

6 FIG. 6 FIG. 1 FIG. 2 2 FIGS.A andB 6 FIG. 6 FIG. 3 FIG. 600 600 101 100 210 256 600 106 600 300 illustrates a second example architecturefor color distortion-aware exposure fusion in accordance with this disclosure. For ease of explanation, the architectureshown inis described as being used by the electronic devicein the network configurationshown inin order to implement at least part of the tone mapping operation(such as the fusion operation) shown in. However, the architecturemay be used with any other suitable device(s) (such as the server) or pipeline(s) and in any other suitable system(s). In, note that thicker lines are used to denote multiple images or image frames and thinner lines are used to denote single images or image frames. The architectureshown inis similar to the architectureshown in, and only the differences between the two are discussed below.

6 FIG. 304 320 602 604 602 302 606 302 302 604 302 608 302 302 a b As shown in, the blending map generation operationand the blending map modification operationhave been replaced with a luma blending map generation operationand a chroma blending map generation operation. The luma blending map generation operationprocesses the multi-exposed image framesand one or more luma tuning parametersto generate initial blending maps for the luma channelsof the multi-exposed image frames. Similarly, the chroma blending map generation operationprocesses the multi-exposed image framesand one or more chroma tuning parametersto generate initial blending maps for the chroma channelsof the multi-exposed image frames.

602 604 304 602 604 302 606 608 608 604 302 th 3 FIG. Both of the blending map generation operationsandmay operate in a similar manner as the blending map generation operationdescribed above. For example, each of the blending map generation operationsandmay determine, for each imulti-exposed image frame, a composite of (i) a first map of values of a contrast or saliency metric C and (ii) a second map of values of a color saturation metric S. These metrics are defined above with respect to. However, the one or more luma tuning parameterscan differ from the one or more chroma tuning parameters. For instance, the one or more chroma tuning parameterscan cause the chroma blending map generation operationto use higher weights with respect to lower-exposure image frameswhen generating the composite values.

7 FIG. 3 6 FIGS.and 7 FIG. 1 FIG. 700 302 300 600 700 604 302 302 700 101 100 700 106 b b illustrates an example techniquefor identifying blending weights for chroma channelsin the architecturesandofin accordance with this disclosure. For example, the techniquecould be used by the chroma blending map generation operationto generate initial blending weight maps for the chroma channelsof the multi-exposed image frames. For ease of explanation, the techniqueshown inis described as being used by the electronic devicein the network configurationshown in. However, the techniquemay be used with any other suitable device(s) (such as the server) and in any other suitable system(s).

7 FIG. 3 FIG. 302 702 706 702 706 302 702 302 302 704 302 302 706 302 302 708 702 706 710 302 302 706 320 320 304 302 706 b As shown in, the multi-exposed image framesare received and processed using multiple weight calculation functions-. The weight calculation functions-are used to generate different initial weights for each of at least some of the image frames. For example, the weight calculation functioncan be used to calculate well-exposedness weights for each image frame, where the well-exposedness weights are based on whether pixels in the image framesare over-exposed, well-exposed, or under-exposed. The weight calculation functioncan be used to calculate saliency weights for each image frame, where the saliency weights are based on contrasts of the pixels in the image frames. The weight calculation functioncan be used to calculate color saturation weights for each image frame, where the color saturation weights are based on levels of color saturation of the pixels in the image frames. A weight combination functioncan combine the weights (such as per pixel) as determined by the weight calculation functions-in order to generate blend weightsfor the chroma channelsof the multi-exposed image frames. Note that at least the weight calculation functioncould be used inas part of the blending map modification operation, where the blending map modification operationmodifies weights of the blending maps from the blending map generation operationbased on color saturations of pixels in the image framesas determined using the weight calculation function.

710 As part of this process, the blend weightscan be determined in a manner that favors the lower-exposure image frames for pixels with saturated colors, reducing contribution by the higher-exposure image frames for those pixels. In some embodiments, to identify pixels with color saturation, RGB pixel values can be analyzed. For instance, pixels with color saturation may often have RGB values in which the largest of the RGB values is much greater than the smallest of the RGB values.

302 302 302 sat As a particular example of this, for each pixel in each image frame, the largest of the pixel's RGB values max(R,G,B) and the smallest of the pixel's RGB values min(R,G,B) can be identified. If the image frameis not in the RGB color space, the image framecan be converted to this color space in order to identify max(R,G,B) and min(R,G,B) for each pixel. The saturation level Lfor each pixel can be calculated as follows.

max min In some embodiments, the functions ƒand ƒcould be defined as follows.

max min max min max min sat sat sat 2 2 8 8 FIGS.A andB 3 6 FIGS.and 8 FIG.A 8 FIG.B 800 802 300 600 706 Here, μand μrespectively represent the means of the maximum and minimum RGB values, and σand σrespectively represent the variances of the maximum and minimum RGB values.illustrate example curvesandfor identifying color saturation weights in the architecturesandofin accordance with this disclosure. More specifically,illustrates an example curve for ƒ, andillustrates an example curve ƒ. When multiple exposures are present, a saturation level at each exposure i can be expressed as L(i) and may be calculated as described above. The saturation levels can be used to calculate a saturation weight W(i), which can represent the output of the weight calculation function. In some cases, the saturation weight W(i) can be expressed as follows.

708 710 In some cases, the weight combination functioncan calculate the blend weightsas follows.

1 2 k k 702 704 Here, W(), W(), . . . represent weights determined by other weight calculation functions-.

3 8 FIGS.through 3 8 FIGS.through 3 8 FIGS.through 5 FIG. 300 600 306 318 322 334 300 600 508 510 Althoughillustrate examples of architecturesandfor color distortion-aware exposure fusion and related details, various changes may be made to. For example, various components, operations, or functions in each ofmay be combined, further subdivided, replicated, omitted, or rearranged and additional components, operations, or functions may be added according to particular needs. Also, while components-and components-are shown as being separate components in the architecturesand, it is possible to use the same components repeatedly (such as serially) to generate the fused luma and fused chroma images. Similarly, while multiple instances of the hue calculation functions-are shown in, the same hue calculation function may be used repeatedly (such as serially) to generate hue maps.

9 9 FIGS.A andB 9 FIG.A 9 FIG.B 900 900 902 904 904 906 906 illustrate example results obtainable using color distortion-aware exposure fusion in accordance with this disclosure. More specifically,illustrates an example output imagegenerated without color distortion-aware exposure fusion. As can be seen here, the output imagehas a large areaof color distortion created as a result of bright illumination from a light source in a scene.illustrates an example output imagegenerated with adaptive multi-stage super-resolution as described above. As can be seen here, the output imagehas a much smaller areaof color change, which is consistent with the bright illumination from the light source. However, there is not a large area of color distortion surrounding the smaller area.

9 9 FIGS.A andB 900 904 Note that other or additional types of benefits can also be obtained using color distortion-aware exposure fusion. For example, if the wall in the background of the scene inincluded a neon sign, the output imagecould have duller or washed-out colors that appear somewhat hazy, which is due to color saturation. However, using the techniques described above, the same neon sign in the output imagewould have more-vibrant colors and appear clearer. This is due to the use of blending weights that favor pixels from lower-exposure image frames over pixels from higher-exposure image frames during blending, since the higher-exposure pixels are more likely to suffer from color saturation.

9 9 FIGS.A andB 9 9 FIGS.A-B 9 9 FIGS.A-B Althoughillustrate one example of results obtainable using color distortion-aware exposure fusion, various changes may be made to. For example,are merely meant to illustrate one example of a type of benefit that might be obtained using the techniques of this disclosure. The specific results that are obtained in any given situation can vary based on the circumstances and based on the specific implementation of the techniques described in this disclosure.

10 FIG. 10 FIG. 1 FIG. 2 2 FIGS.A andB 3 8 FIGS.through 1000 1000 101 100 101 200 300 600 106 illustrates an example methodfor color distortion-aware exposure fusion in accordance with this disclosure. For ease of explanation, the methodshown inis described as being performed by the electronic devicein the network configurationshown in, where the electronic devicecan implement the pipelineshown inand include one of the architecturesandshown in. However, the method may be performed using any other suitable device(s) (such as the server), pipeline(s), or architecture(s) and in any other suitable system(s).

10 FIG. 1002 120 101 302 302 252 250 254 302 As shown in, multiple image frames of a scene are obtained at step. This may include, for example, the processorof the electronic devicegenerating or otherwise obtaining multiple multi-exposed image frames. In some cases, the multiple multi-exposed image framesmay be generated by performing one or more synthetic exposure generation operationsusing a multi-exposed image, such as a linear or other HDR image, to generate multiple synthetic-exposure image frames. Each multi-exposed image frameincludes a luma channel and chroma channels.

1004 120 101 302 300 600 1006 120 101 302 303 302 302 a b Blending of the image frames to generate a blended image is initiated at step. This may include, for example, the processorof the electronic deviceinitiating processing of the multi-exposed image framesusing the architectureor. As part of the blending process, weights to be applied when blending the luma and chroma channels of the image frames are determined at step. This may include, for example, the processorof the electronic devicedetermining first weights to be applied when blending the luma channelsof the image framesand separate second weights to be applied when blending the chroma channelsof the image frames.

302 404 402 406 404 404 302 302 302 302 302 302 302 514 302 302 302 514 b b As noted above, there are various ways in which the first and second weights can be determined. In some embodiments, the first weights may be determined based on the image frames, and the first weights may be modified to generate the second weights. As a particular example, the first weights can be modified by applying an exponential or power-law functionto weights of original blending mapsin order to produce modified blending maps. The exponential or power-law functioncan reduce first ones of the first weights and/or increase second ones of the first weights, where the first ones of the first weights have smaller values than the second ones of the first weights. The exponential or power-law functioncauses the second weights, compared to the first weights, to favor the chroma channelsof one or more first ones of the image framesover the chroma channelsof one or more second ones of the image framesduring the blending of the image frames, where the one or more first ones of the image frameshaving a lower exposure than the one or more second ones of the image frames. As another particular example, the first weights can be modified by determining error mapsbased on hues of pairs of the image frames(where one of the two image frameshas a lower exposure than another of the two image frames), and the first weights can be modified based on the error maps.

606 608 608 606 302 302 702 706 302 708 710 302 302 302 302 302 In other embodiments, the first weights and the second weights can be determined separately, such as by determining the first weights using one or more luma tuning parametersand separately determining the second weights using one or more chroma tuning parameters. The one or more chroma tuning parameterscan be different than the one or more luma tuning parameters. In some cases, the second weights can be generated by determining different initial weights for the image frames, where the different initial weights are associated with different characteristics of the image frames. For instance, the weight calculation functions-can be used to generate the initial weights of the image frames, such as initial weights based on well-exposedness, saliency, and color saturation. The different initial weights for the image frames can be combined to generate the second weights, such as by using the weight combination functionto calculate blend weights. In some cases, the different initial weights for the image framescan include color saturation weights associated with the image frames, where the color saturation weights are based on amounts of color saturation within the image frames. When combining the initial weights, certain ones of the initial weights that are associated with larger amounts of color saturation within the image framescan be reduced, and/or other certain ones of the initial weights that are associated with smaller amounts of color saturation within the image framescan be increased.

302 302 302 302 302 302 302 302 b b In various embodiments, the image framesinclude at least one lower-exposure image frame and at least one higher-exposure image frame. The second weights can be determined so that, during the blending of the image frames, (i) the contribution of the chroma channelsof the at least one lower-exposure image frameis increased and/or (ii) the contribution of the chroma channelsof the at least one higher-exposure image frameis decreased. This may, for example, help to reduce or avoid issues with color saturation since the at least one lower-exposure image framelikely has less over-saturation compared to the at least one higher-exposure image frame.

1008 120 101 302 302 120 101 302 302 1010 120 101 212 338 a b A fused luma image and at least one fused chroma image are generated at step. This may include, for example, the processorof the electronic devicegenerating the fused luma image using the luma channelsof the image framesand the first weights. This may also include the processorof the electronic deviceseparately generating the at least one fused chroma image using the chroma channelsof the image framesand the second weights. The fused luma image and the at least one fused chroma image are combined to generate the blended image at step. This may include, for example, the processorof the electronic devicecombining the fused luma image and the at least one fused chroma image into a single output image,.

1012 212 338 160 101 130 101 101 212 338 160 101 130 101 101 212 338 The blended image can be stored, output, or used in some manner at step. For example, the output image,may be displayed on the displayof the electronic device, saved to a camera roll stored in a memoryof the electronic device, or attached to a text message, email, or other communication to be transmitted from the electronic device. As another example, multiple output images,could be generated and used to form a video sequence, which again can be displayed on the displayof the electronic device, saved to the camera roll stored in the memoryof the electronic device, or attached to the text message, email, or other communication to be transmitted from the electronic device. Of course, the output image,could be used in any other or additional manner.

10 FIG. 10 FIG. 10 FIG. 1000 Althoughillustrates one example of a methodfor color distortion-aware exposure fusion, various changes may be made to. For example, while shown as a series of steps, various steps inmay overlap, occur in parallel, occur in a different order, or occur any number of times (including zero times).

2 10 FIGS.through 2 10 FIGS.through 2 10 FIGS.through 2 10 FIGS.through 2 10 FIGS.through 101 102 104 106 120 101 102 104 106 It should be noted that the functions shown in or described with respect tocan be implemented in an electronic device,,, server, or other device(s) in any suitable manner. For example, in some embodiments, at least some of the functions shown in or described with respect tocan be implemented or supported using one or more software applications or other software instructions that are executed by the processorof the electronic device,,, server, or other device(s). In other embodiments, at least some of the functions shown in or described with respect tocan be implemented or supported using dedicated hardware components. In general, the functions shown in or described with respect tocan be performed using any suitable hardware or any suitable combination of hardware and software/firmware instructions. Also, the functions shown in or described with respect tocan be performed by a single device or by multiple devices.

Although this disclosure has been described with example embodiments, various changes and modifications may be suggested to one skilled in the art. It is intended that this disclosure encompass such changes and modifications as fall within the scope of the appended claims.